Cell fate specification during development is often thought of as a highly reproducible process driven by cell lineage and signaling. Cellular diversity can also arise from the inherent molecular noise (i.e. variability) in gene expression during stochastic cell fate specification, where a cell randomly chooses between two or more fates. Compared to lineage and signaling mechanisms, very little is known about how noise in gene expression drives fate decisions during development. This project aims to address stochastic cell fate specification in the visual system of Drosophila melanogaster. The fly eye contains a simple, stochastic, binary fate choice. The eye is a random mosaic of two color-detecting photoreceptor subtypes, defined by expression of different light-detecting Rhodopsin proteins. This binary decision is controlled by the transcription factor Spineless (Ss), which is expressed in a random subset (65%) of R7 photoreceptors. Stochastic expression of the gene spineless (ss) is controlled in a temporal manner throughout development by two enhancer elements (?early? and ?late? enhancers), and a transcriptional repressor, Klu. Early expression of ss is noisy, producing variability in the levels of gene expression between precursor R7 cells. By manipulating DNA elements (enhancers and silencers) and the Klu transcription factor, the strength of ss expression early can be tuned. Remarkably, these manipulations of early expression cause changes in the on/off ratio of Ss in terminally differentiated R7 cells. Based on these findings, it is hypothesized that variable levels of ss expression amongst R7 progenitors (i.e. early) sets the on/off expression frequency in terminally differentiated R7 cells (i.e. late). This project aims to address what promotes early ss expression variability as well as the consequences of this variability. A major source of gene expression noise arises from the process of transcription. Transcription is inherently stochastic, and occurs in bursts that vary in amplitude, frequency, and duration between genes and cells. Three-color RNA fluorescence in situ hybridization will be used to monitor transcription in individual cells to determine how Klu and different DNA elements control variation in in early ss expression in fixed tissue (Aim 1). To gain a mechanistic understanding of how transcriptional affects early expression variability, this project aims to visualize early ss expression by using the MS2/MCP system and live imaging (Aim 1). Finally, a link between early expression and terminal cell fate will be made by monitoring endogenous ss expression in real time from the precursor cell stage through terminal differentiation (Aim 2). This project will be carried out at Johns Hopkins University in the lab of Robert J. Johnston Jr. The applicant will receive additional training from collaborators at Princeton University and the Institut Pasteur (Dr. Thomas Gregor). The results of this project will elucidate the mechanisms controlling expression of ss during retina development, in particular, how molecular variability drives a stochastic cell fate decision during metazoan development